Shall We Stop ChatGPT From Visiting the Colony?

[AgentPete]

You may have seen that ChatGPT intends to scour the net, collecting data for the next iteration of its Large language Model (LLM). It’s basically going to be slurping up everything it can find… heaven knows how they’re going to store all that data.

We have the option here of blocking ChatGPT’s bot from accessing the Colony and I wanted to know your thoughts about that.

As you know, certain areas of the Colony are not visible to Google and other search engines (The Back Room & the Laboratory). Parts of other areas are indeed search-engine-visible, because we need to be indexed by Google so that folk can find us.

I think the same is partly true of ChatGPT, although whether being in their LLM really will add to discoverability is, frankly, anyone’s guess at the moment.

My own feeling is that we should probably block their bot from visiting here, but I’d like your thoughts and vote in the poll, please. Poll will close on Saturday.

 

[Serra K]

One vote for blocking.

 

[Pamela Jo]

Is there a hell no fuck the horse it rode in on option?

 

[Rich.]

I think Chat GPT, and AI in general, is fascinating and exciting, but the tech (yet again) seems so far ahead of any kind of ethically-informed regulatory framework that I'd be inclined to wait before making Litopia's data available to it. So that's a block vote from me as well.

[edited for clarity in light of @Emily's comment below]

 

[Emily]

I vote to block it.... but just to note @AgentPete that while people are saying No to disallowing it... the "voting No" above is actually allowing it, and the "Voting Yes" is to block it (if that makes sense )

 

[Bloo]

...we need to be indexed by Google so that folk can find us.

Soo...what's stopping the Googolplex or Bing from bolting LLM scraping functionality onto indexing bots? Allow LLM or lose discovery. Helluva choice.

A less "restrictive" and perhaps more insideous algorithm could push sites further down search results if they don't play ball.

Maybe even charge a fee to disallow LLM scraping while still being indexed?

There are so many ways this can go down.

@AgentPete - in case the future really gets this bleak, would limiting index bots to the home page significantly impact discovery?

BTW...I voted for the ole blockeroo.

 

[AgentPete]

Soo...what's stopping the Googolplex or Bing from bolting LLM scraping functionality onto indexing bots? Allow LLM or lose discovery. Helluva choice.

A less "restrictive" and perhaps more insideous algorithm could push sites further down search results if they don't play ball.

Maybe even charge a fee to disallow LLM scraping while still being indexed?

There are so many ways this can go down.

Well quite.

I think OpenAI want to be seen to be behaving responsibly here, so they’re giving their bot a name and say it will obey robots.txt directives, but there are many other ways to scrape a website, of course. As you say, OpenAI’s friends / beastmasters could easily push them vast quantities of scraped data. Or they could simply do what some other nefarious bots do, which is to ignore robots.txt.

@AgentPete - in case the future really gets this bleak, would limiting index bots to the home page significantly impact discovery?

Yes, this could be done, but not by robots.txt which is a convention and not a firewall. Locking the entire site to be viewable by members-only (apart from the home page) would be one option. Would demote us in search results, tho.

 

[AgentPete]

I vote to block it.... but just to note @AgentPete that while people are saying No to disallowing it... the "voting No" above is actually allowing it, and the "Voting Yes" is to block it (if that makes sense )

So logical
Now clarified!

 

[Hannah F]

Block it, I say, block it.

Vote casted.

 

[James Charles]

Voting, but it will find a wormhole to wiggle its way in.

 

[Bloo]

Voting, but it will find a wormhole to wiggle its way in.

I think it already found the poll. There's one "don't block" vote.

 

[CarolMS]

Block it, yes. We have to do what we can to maintain some semblance of safety--and sanity. Unfortunately, it's the latter I worry about, especially in the form of narcissistic self-righteousnous and hubris. And greed. I don't understand any of this tech stuff, but I do understand human nature. However many fine, flawless apples are in the basket, there are always a few rotten ones, or at least some bruised beyond recognition. Just imagine how many folks there are out there like the guy who overly hyped and stupidly believed in his Titan submersible. No doubt many of them are working on AI, and god knows whatever other "revolutionary" technology coming down the pike. Sorry to be so pessimistic. I just think it can't be kept up with, and the monster is out of the bottle and can't be put back in. Sure there have always been similar scenarios that we've overcome or learned to control, but we've never seen anything like what is being created today and will be tomorrow. I'm not sure precedent applies; that seems naive. Still, in our small ways, we can do what we can. What else is there? (Just my two cents worth and not intended as a conversation thread--at least not one I can participate in, since I'm finding it hard to do so due to my circumstances and limitations these days.)

 

[Rachel Caldecott]

I think Chat GPT, and AI in general, is fascinating and exciting, but the tech (yet again) seems so far ahead of any kind of ethically-informed regulatory framework that I'd be inclined to wait before making Litopia's data available to it. So that's a block vote from me as well.

[edited for clarity in light of @Emily's comment below]

Hmm, yes it is fascinating, but the "ethically-informed regulatory framework" that you dream of is a) in human hands, (and several thousand years of human history might suggest that we can't be trusted on that score) and b) is going to be wrenched from human hands by AI.

So, I say we need to hold it at bay for as long as possible. No matter how much it fascinates and excites us.

I've said it before, and I'll say it again. We are at a crisis point. On the one hand, we have max 10 years before climate breakdown changes society into something unrecognisable (and probably highly unpleasant). On the other, we are on the cusp of creating a technology-run economy that will cause untold suffering. We are literally standing between the Hollywood movies, The Day After Tomorrow, and I, Robot.

 

[Rich.]

Hmm, yes it is fascinating, but the "ethically-informed regulatory framework" that you dream of [...]

I don't think I said I was dreaming of it.

As to the future of AI, it's certainly a subject that provokes a range of strong opinions!

I don't know. Labour laws, for example, in Europe are certainly better than they were a century ago. But equally, civilizations fall (you can take it as read, @Rachel Caldecott, that I agree with you on climate change). I'm less convinced by the arguments of sentient AI taking over the world. But I genuinely don't have strong opinions on this one (for want of deep enough knowledge): once you get away from the sci-fi informed predictions, the jury does seem to be out on whether AI could, even in principle, become self aware.

But, coming back to the topic of this thread and as reflected in my vote, I don't think we should rush to unconditionally give up our data to commercial experiments.

 

[Rachel Caldecott]

I think Chat GPT, and AI in general, is fascinating and exciting, but the tech (yet again) seems so far ahead of any kind of ethically-informed regulatory framework that I'd be inclined to wait before making Litopia's data available to it. So that's a block vote from me as well.

[edited for clarity in light of @Emily's comment below]

I completely misread your original comment and thought you were all for letting AI in to run through Litopia's corridors. My bad!

 

[Bloo]

I completely misread your original comment and thought you were all for letting AI in to run through Litopia's corridors. My bad!

Sticks out foot as ChatGPT runs past.

 

[Rich.]

I completely misread your original comment and thought you were all for letting AI in to run through Litopia's corridors. My bad!

No worries at all! *hugs*

 

[izi 出久]

I think Chat GPT, and AI in general, is fascinating and exciting, but the tech (yet again) seems so far ahead of any kind of ethically-informed regulatory framework that I'd be inclined to wait before making Litopia's data available to it. So that's a block vote from me as well.

[edited for clarity in light of @Emily's comment below]

I second this. I’m not in any way concerned about AIs in general. But we need regulation before setting this baby free. This is a block vote

 

[AgentPete]

I think it already found the poll. There's one "don't block" vote.

Yeah, the verdict is pretty overwhelming

 

[Ancora Imparo]

Count me as another block vote. And if you can give it a good kicking as you show it the door, even better.

 

[Eva Ulian]

I'm in favour of blocking it too. That's not because I am against AI as such, but I get a creepy feeling thinking someone is watching every word I type on this forum, storing it and using it in heaven knows what way.

 

[AgentPete]

OK, block now in place. We're not the only ones

 

[Jonny]

Dang! I forgot to vote but it would've been a yes too.

 

[Hannah F]

Count me as another block vote. And if you can give it a good kicking as you show it the door, even better.